Reducing the feedback overhead during crosstalk precoder initialization
An apparatus comprising a receiver coupled to a digital subscriber line (DSL) between an exchange site and a customer premise equipment (CPE) and configured to send a feedback error message to train a precoder coupled to the exchange site, wherein the feedback error message comprises a plurality of error components and an indication of a quantity of bits per error component, a quantization accuracy per error component, or both. Included is a method comprising sending an error feedback message to a DSL crosstalk precoder to train the crosstalk precoder, wherein the error feedback message comprises an error vector and a quantization scaling factor of the error vector.
1. A digital subscriber line (DSL) transceiver of a customer premise equipment (CPE), comprising:
a memory storing instructions; and
a processor coupled to the memory and executing the instructions to:
measure a pilot signal and generate a plurality of subcarrier error measurements to be reported, with the pilot signal being received over a DSL link from a DSL system and with a subcarrier error measurement of the plurality of subcarrier error measurements including a real component and an imaginary component;
determine a quantization factor based on a maximum magnitude subcarrier error measurement in the plurality of subcarrier error measurements;
transfer an error measurement portion from each subcarrier error measurement of the plurality of subcarrier error measurements into a corresponding first field of a feedback error message, with each error measurement portion selected using the quantization factor and wherein the feedback error message comprises a plurality of first fields for carrying the plurality of error measurement portions and a second field for carrying the quantization factor; and
insert the quantization factor into the second field of the feedback error message, wherein the DSL transceiver is configured to transmit the feedback error message to a precoder of the DSL system over the DSL link, and wherein the feedback error message is used for precoder training.
2. The DSL transceiver of claim 1 , wherein the quantization factor is determined based on the maximum magnitude subcarrier error measurement and a quantization accuracy.
3. The DSL transceiver of claim 1 , wherein the quantization factor comprises one or both of a quantity of bits in the maximum magnitude subcarrier error measurement and a quantization accuracy.
4. The DSL transceiver of claim 3 , wherein the quantization accuracy is fixed and the quantity of bits is determined by:
N
r
(
i
)
=
log
2
(
2
e
max
(
i
)
d
)
,
where N r (i) is the quantity of bits for the sync symbol i and d is the quantization accuracy.
5. The DSL transceiver of claim 3 , wherein the quantity of bits is fixed and the quantization accuracy is determined by:
d
(
i
)
=
2
e
max
(
i
)
2
N
r
,
where N r (i) is the quantity of bits for the sync symbol i and d is the quantization accuracy.
6. The DSL transceiver of claim 1 , wherein an error measurement portion corresponds to a level of crosstalk error in the DSL link.
7. The DSL transceiver of claim 1 , wherein the pilot signal comprises a sequence of pilot symbols and each pilot symbol of the sequence of pilot symbols comprises a plurality of tones, wherein the error measurement comprises an error measurement vector represented in complex format:
E ( k,i )= e x ( k,i )+ j·e y ( k,i ),
where E(k,i) is an error measurement vector for a sync symbol i and a tone k of the sync symbol, e x (k,i) is a real component of the error measurement, and e y (k,i) is an imaginary component of the error measurement.
8. The DSL transceiver of claim 7 , wherein the maximum magnitude subcarrier error measurement is obtained from the real component e x (k,i) and the imaginary component e y (k,i) by:
e
max
(
i
)
=
max
k
{
max
{
e
x
(
k
,
i
)
,
e
y
(
k
,
i
)
}
}
,
where e max (i) is a maximum error in the sync symbol i and max{ } indicates a function for selecting a maximum sample from a set.
9. The DSL transceiver of claim 7 , wherein the error measurement vector is scaled using a scaling factor:
ē x ( k,i )= S Q ( i )· e x ( k,i ),
ē y ( k,i )= S Q ( i )· e y ( k,i ),
where S Q (i) is the scaling factor, ē x is a scaled error measurement vector for the real component of the error measurement and ē y is a scaled error measurement vector for the imaginary component of the error measurement.
10. The DSL transceiver of claim 7 , wherein the error measurement represented using a quantization format:
E x ( k,i )=max{−2 N-1 ,min{ S Q ( i )· e x ( k,i )·2 N-1 ,2 N-1 −1}},
E y ( k,i )=max{−2 N-1 ,min{ S Q ( i )· e y ( k,i )·2 N-1 ,2 N-1 −1}},
where E x (k,i) is the real component of the error measurement and E y (k,i) is the imaginary component of the error measurement.